Sierkinhane CRNN_Chinese_Characters_Rec issues

Sierkinhane / CRNN_Chinese_Characters_Rec

(CRNN) Chinese Characters Recognition.

1.81k stars 537 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

训练期间得到的结果很好，但是测试（即使使用训练集的图片）只数出第一个字符，请问这是为什么

#269 Matthew-dhu opened 3 years ago
11
不小心ctrl+c中断了进程，继续训练的话，RESUME那改成了中断epoch前保存的权重，但是学习率是不是没有保存，会不会对训练有影响。

#268 chichuhu opened 3 years ago
0
解决了单机多卡并行训练的问题

#267 chichuhu opened 3 years ago
2
mode.(inp).cpu()

#266 Donaghys closed 3 years ago
0
单机多卡训练损失不下降

#265 Donaghys closed 3 years ago
0
模型输出是nan

#264 dengfenglai321 opened 3 years ago
2
训练开始loss就是nan

#263 dengfenglai321 closed 3 years ago
5
demo 自訓練模型問題

#262 yuanhuang0825 closed 3 years ago
0
单机多卡

#261 Donaghys closed 3 years ago
2
关于训练自己数据集的问题

#260 dengfenglai321 closed 3 years ago
2
KeyError: 'ｄ'

#259 dengfenglai321 closed 3 years ago
0
有大佬给出不定长图片和标签训练应改哪地方的代码吗，实在没思路了

#258 chichuhu opened 3 years ago
6
訓練速度越來越慢

#257 yuanhuang0825 opened 3 years ago
1
能训练手写体不定长识别吗

#256 chichuhu opened 3 years ago
3
训练时候一直无法预测导致accuray一直为0

#255 yuanyuan1993 opened 3 years ago
17
请问不能直接拿字典文件生成识别数据吗？为什么还得需要语料

#254 Donaghys opened 3 years ago
1
train 的時候輸出都是空白

#253 yuanhuang0825 opened 3 years ago
5
中文姓名识别的建议

#252 Ryansanity opened 3 years ago
0
您好请问训练的时候损失正常下降，但是验证集的预测为空，这是为什么

#251 Donaghys closed 3 years ago
1
在服务器上运行train.py 一直无法解决这个问题是cv的版本不对吗可以告知下详细的环境配置吗

#250 Agiroy4712 opened 3 years ago
5
想问一下大家，训练第一个epoch的准确是多少，总感觉自己的有点不正常

#249 zhangchunyang1995 opened 3 years ago
2
Synthetic Chinese String Dataset数据集训练速度

#248 whwme opened 3 years ago
6
why you use cpu --->model(inp).cpu() ...............................

#247 cqray1990 opened 4 years ago
3
请问下这个cnn用的是哪个backbone?

#246 oftenliu opened 4 years ago
1
训练数据和测试数据长度分布问题

#245 wolfryu opened 4 years ago
2
请问300W数据，训练一个epoch以后Test loss: 0.0086, accuray: 0.9872，正常吗？

#244 zhiaoLee opened 4 years ago
1
train的时候数据准确率上不去怎么解决?

#243 linyilong3 opened 4 years ago
0
自己训练数据acc大于1，效果不理想

#242 zhiaoLee opened 4 years ago
1
训练速度过慢，且gpu利用率很低

#241 first-wording opened 4 years ago
1
训练不同长度标签

#240 Springzcf opened 4 years ago
18
char_std_5990.txt 文件乱码

#239 WJinhai closed 4 years ago
1
迁移学习训练后只能识别一种类型

#238 hfdzlsw opened 4 years ago
4
请问标注工具是什么，或者有什么推介的吗？

#237 linyilong3 opened 4 years ago
1
请问为什么要对input image进行resize？

#236 kylechang523 closed 4 years ago
1
index = self.dict[char] KeyError: '闈'

#235 lucasjinreal closed 4 years ago
6
中断续训练的问题

#234 first-wording closed 4 years ago
2
AttributeError: 'EasyDict' object has no attribute 'FINETUNE'

#233 first-wording closed 4 years ago
2
训练高度为64的图片会有问题

#232 PKQ1688 opened 4 years ago
12
txt文件格式

#231 MoEmma closed 4 years ago
2
How to train model with linear layer or conv layer?

#230 feitiandemiaomi opened 4 years ago
4
加载train.py时出错

#229 freedom-xu-ming-xue closed 4 years ago
2
自己训练数据acc达到90%左右，测试图片可以识别，新的图片识别效果很差，不清楚为什么

#228 18796956836 opened 4 years ago
1
模型微调

#227 Lie-huo opened 4 years ago
2
请问训练后输出的模型在哪里？

#226 qingzhu0214 closed 4 years ago
3
推理精度

#225 Zrufy closed 4 years ago
1
大佬，我使用gpu训练的过程中，运行一小会，就变成cpu在训练了，并且loss=nan，各位有遇到的吗？

#224 jiaoxiaosong opened 4 years ago
0
自己数据集（3W）如下。怎么才能提高准确率

#223 XueKX closed 3 years ago
11
GPU上无法运行

#222 a-whitej opened 4 years ago
2
360w数据集解压的时候损坏，使用好压修复也不行

#221 jiaoxiaosong closed 4 years ago
2
精度及其低

#220 liu-ca opened 4 years ago
1

Previous Next